Forecasting and Reservation of Transcoding Resources for Live Streaming

2. The system of claim 1, wherein the plurality of cloud sites are included in a geo-distributed cloud infrastructure.

3. The system of claim 1, wherein the video information includes information on a plurality of videos broadcasted at the plurality of cloud sites.

4. The system of claim 1, wherein the video information includes location information of the viewers of a video and a broadcasted video quality of the video.

5. The system of claim 1, wherein each respective cloud site includes its own respective collected video information.

6. The system of claim 1, wherein the at least one model is trained by one or more machine learning algorithms including Long Short Term Memory (LSTM), Gated Recurrent Unit (GRU), Convolutional Neural Network (CNN), MultiLayer Perceptron (MLP), and XGboost.

8. The method of claim 7, wherein the decision variable A(v, ra) is equal to one if the incoming video v is allocated in the video allocation region ra and is equal to zero otherwise.

11. The method of claim 10, wherein the decision variable W(v, ra, rw) is equal to one if viewers in the video serving region rw are served from region ra and is equal to zero otherwise.

12. The method of claim 7, further comprising determining from which respective cloud site of the plurality of cloud sites that a video should be broadcast to one or more viewers.

13. The method of claim 12, wherein the respective cloud site from which the video should be broadcast is the cloud site that is nearest geographically to the one or more viewers.

15. The method of claim 14, further comprising training the at least one model using the received desired quantity of cloud instances.

16. The method of claim 14, further comprising updating the video information with the received desired quantity of cloud instances.

17. The method of claim 14, wherein the video information includes one or more of a length, size, bitrate, frame rate, resolution, and one or more geographical locations of viewers of the one or more videos streamed at a first time.